EXTRACTED FROM THE AMERICAN JOURNAL OF SCIENCE AND ARTS, VOL. Ill, 
SECOND SERIES, 1847. 
ORIGIN 
OF THE 
GRAND OUTLINE FEATURES 
OF THE 
EARTH. 
BY 
JAMES D. DANA. 
The traveller who follows along the indented shores of a con- 
tinent, or, traversing the ocean, touches here and there at an 
island, is likely to become strongly impressed with the common 
idea that the land has every where the utmost irregularity of 
form, and that islands are “ scattered dots” over the wide seas. 
Such must be the conclusion of any one who should judge from 
his own casual observations. Were there a system in the world’s 
physiognomy, it could not become apparent till explorations had 
traced accurately the varying outlines and elevations of continents, 
and the positions of islands. A good map or globe is a register 
of the thousand observations of voyagers and surveyors upon 
these points. It is a miniature of the world within the grasp of 
Amer. Jour. Sci., ii Ser., Vol. III.—May, 1847. 1 382 
Origin of the Grand Outline Features of the Earth. 
our view, and may be properly appealed to with reference to the 
earth’s features. The evidence has not been overlooked by the 
philosophers of the day, and is more or less fully discussed by 
Humboldt, Malte Brun, L. A. Necker, Elie de Beaumont, Boue, 
and other geographers and geologists. Yet it has failed of fixing 
general attention. It is proposed to pass briefly in review the 
principal facts, and consider the causes to which the existing fea- 
tures of the earth are attributable. 
The remarks which follow will be hardly intelligible to the 
reader without a globe before him, or a Mercator’s chart of the 
world: and the latter, though the best kind of map for the pur- 
pose, is somewhat erroneous in consequence of the parallelism of 
the meridians. 
Trends of Coasts and other Features of Continents.—1. On the 
continent of America, the reader will observe the nearly straight 
coast line from the Gulf of Mexico along by Newfoundland and 
Greenland, a distance of 5000 miles ; the near parallelism of this 
line with the southeast coast of South America, 4000 miles in 
length; also with the line of Lakes Ontario and Erie, and the river 
St. Lawrence; also with the coast on the northwest of Hudson’s 
Bay, and that by Prince Regent’s inlet. These parallelisms are 
too striking not to be at once obvious. They are instances of a 
northeast and southivest trend; and the distance between these 
great parallel lines are respectively about 3000 miles, 250 miles, 
1400, and 380 miles. 
Again : look at the west coast of the same continent, from Da- 
rien to Russian America, and laying down a rule, mark the near 
parallelism of the course with the line of great lakes, from Erie 
through Michigan, Superior, Winnipeg, Slave and Bear, to the 
coast by the mouth of the Mackenzie ; also with the southwest 
side of Hudson Bay ; with the coast on the west of Davis 
Strait and Baffin Bay, and that also on the east. Here the 
uniformity is even more remarkable. These are instances of a 
northwest and southeast trend. To one of the two lines corres- 
pond the greater part of all the grand features of the continent. 
The apparent exceptions will be hereafter considered. The dis- 
tances separating these northwest lines average respectively 1000 
miles, 350, 700, and 400 miles. 
2. Compare the sides of the Atlantic Ocean. The southeast 
coast of South America, from Magellan to St. Roque, is almost 
an exact continuation of the western line of the opposite conti- 
nent, by western Africa, Spain, and Norway or the Baltic ; and 
the break in the line made by the Atlantic, is partly filled by the 
islands Fernando Noronha, St. Paul and the Cape Verds. It will 
be shown hereafter that the northeast coast of South America 
belongs to the northwest system of trends, and extends by 
Guatemala and California northward; and if we cross the ocean, Origin of the Grand Outline Features of the Earth. 
383 
we find the Cape Palmas coast of Africa, or rather the Kong 
mountains adjoining, and the Pyrenees between Spain and France, 
having the same trend as the northeast of South America. 
3. In the eastern continent, the western coast of Europe and 
eastern of Africa have a striking parallelism, in which the north 
coast of Asia, from the Obo Gulf to the northeast cape, partakes; 
and so also the east coast of Asia, the east coast of Hindostan, 
and also the island of Madagascar. These are northeast trends. 
Again, we observe that the Red Sea, Adriatic and British isles, 
the Persian Gulf, Western Hindostan, and the coast from Calcutta 
by Malacca, are nearly parallel lines, having the northwest trend. 
Whatever may be said of the exceptions and irregularities, there 
are evidently too many coincidences to be set aside as mere acci- 
dent ; and the two courses have accordingly been considered by 
Humboldt, the great geographical lines of the globe. 
We might pursue the same course with the mountains. But 
the general parallelism of the chains with the coast lines, is so 
obvious that we barely allude to them in this brief outline of the 
subject under discussion. The fact is plainly true with respect 
to the great chains of America: and some particulars, soon to be 
stated, will bring into one system what appears irregular in Europe 
and Asia. We observe only that Tchihatchef, a Russian trav- 
eller, has pointed out the conformity of the main range of the 
Altai to the course of the east coast of Africa and the island of 
Madagascar, while there is also a northeast range in the same 
chain, parallel with the Persian Gulf and Red Sea: the highest 
point of the Altai mountains occurs at the intersection of the 
two ranges.* 
Trends of Groups of Is lands.—We proceed to a few remarks 
on the islands of the ocean; and they are the more satisfactory, 
as islands are the culminant peaks of submerged mountain chains, 
and therefore mark correctly the proper outline features of the 
earth’s surface. The fact that the islands range in lines was long 
since pointed out; and Malte Brun remarks on the regularity of 
these lines in the Pacific Ocean. Compare the line in which lie 
New Ireland, the Salomon group and the New Hebrides, with 
the Sandwich or Hawaiian range, extending to the northwest 
through a number of small islets 2000 miles, from longitude 155° 
W., beyond 180°. Though 3500 miles apart, the lines are nearly 
parallel; and parallel also with New Caledonia, and with northeast 
New Holland, on one side, and nearly also with the coast of Cal- 
ifornia and Guatimala on the other ; and moreover, these lines are 
* Comptes Rendus, May 12, 1845. Tchihatchef observes that the occidental 
chain runs N. \v. and s. e., and the other or oriental, N. e. and s. w. He mentions 
that a similar system exists in the Alps; but as the identity of direction is not ex- 
act, he considers the two systems different. 384 
Origin of the Grand Outline Features of the Earth. 
parallel with the several intermediate chains of islands, as is appa- 
rent on the largest charts : the details which are highly interesting 
will be given by the writer in his Geological Report on the Pacific. 
Here, then, is an approximate parallelism in ranges over more than 
a quarter of the circumference of the globe; and nearly all cor- 
respond to the northwest trend. New Zealand and the Tonga 
group, and the Ladrones, are examples of the transverse or noi'th- 
easterly trend. The latter trend, (approaching n. n. e.,) character- 
izes a part of New Holland, the Australian Alps of the southeast, 
and as observed by Fitton,# the west coast of the Gulf of Car- 
pentaria and the islands off its north cape ; besides, also, the north- 
west coast: the inlets of the coast correspond almost uniformly 
to one of the two courses pointed out. The Galapagos illustrate 
distinctly both trends, and they are so mentioned without allud- 
ing to any general law, by Darwin.f Beyond the American con- 
tinent, in the Atlantic, we find the Azores closely parallel with the 
Hawaiian line; and the same is illustrated in the Canaries, ac- 
cording to the position of the islands given by von Buch,| and 
also in the Cape Yerds. These lines are also parallel with north- 
east South America, and the Pyrenees. Thus, not only over one 
ocean, but over both, the same system prevails, and alike also 
over the intermediate continents, the one corroborating the other. 
The system in truth belongs to the world. To this conclusion 
Humboldt, Necker, Boue. and other geologists appear to have ar- 
rived. 
But if we survey the facts more minutely, may we not find 
that an element in this branch of physical geography has not 
been properly apprehended ? Do not the exceptions throw in a 
vexatious doubt, if they cannot be blended with the theory? 
We propose, then, to pursue the subject still farther; and we 
believe that instead of proving that there are as many distinct 
systems in orography as there are mountain courses,$ it will 
* Sketch of the Geology of Australia, Phil. Mag., 1826, lxviii, 18,132. The reg- 
ular lines of islands were observed by Flinders, as remarked by Fition, (ibid., p. 
132.) See Flinders’s Voyage, ii, 246. 
1 Volcanic Islands, p. 115. “ Three great craters on Albemarle island, form a 
well marked line, extending n. w. by N. and s. e. by s.; Narborough island and 
the great crater on the rectangular projection of Albemarle island, form a second 
Sarallel line. To the east, Hood’s island and the islands and rocks between it and 
ames’ island, form another nearly parallel line; which, when prolonged, includes 
Culpepper and Wenman islands, lying sevenly miles to the north. The other 
islands, lying further eastward, form a less regular fourth line. Several of these 
islands and the vents on Albemarle island, are so placed that they likewise fall on 
a set of rudely parallel lines, intersecting the former lines at right angles ; so that 
the principal craters appear to lie on the points where two sets of fissures cross 
each other.” 
t Les lies Canaries, 369. The craters of Gran Canary, Teneriffe and Palma, 
are in a northwest line, while a transverse trend is distinct in the several islands. 
§ Essentially the theory of Elie de Beaumont, in which view he is supported by 
many distinguished names in geology. Origin of the Grand Outline Features of the Earth. 
385 
appear more in accordance with facts, to refer even wide devia- 
tions of directions to one and the same system. 
General character of the lines of Mountains, Coasts, and Isl- 
ands.—A careful study of the courses of island groups, coasts 
and mountain chains, leads us to the following important re- 
sults :— 
I. The ranges are made up of shorter consecutive and some- 
times parallel lines, instead of being uninterrupted for long dis- 
tances. 
XI. The ranges are more commonly curved, than straight or 
coincident with the course of a great circle. 
III. The straight ranges are generally straight in the con- 
stituent lines, but may coiisist of a series of curves. 
IV. Curved ranges may arise from a general curvature in the 
whole; but often proceed from the positions of the several con- 
secutive parts. 
V. The same range, owing to the mode of curving, may vary 
greatly in its course, in different portions. 
In these points we are stating mere- 
ly the facts or results of observation, 
free from speculation. The following 
figures may serve to illustrate the prop- 
ositions stated. 
In figure 1, the entire range is 
straight, as well as the parts. 
In figures 2, 3, and 4, the parts are 
straight and overlap, and thus form a 
range which is sometimes straight as 
a whole, but is more frequently curved. 
The direction of the whole range, as 
shown by the dotted line, differs from 
the direction of the subordinate lines. 
Figures 3 and 7 represent a common condition in which there 
are parallel lines in some parts of a range. 
In figure 5, the parts are curved; and here, too, the resulting 
range may be straight or curved. 
Fig. 6 represents a range made up of longitudinal parts along 
with some transverse ; this is of common occurrence (figure 7.) 
The more thorough the examination of the trends of groups of 
islands and of mountain chains, the more distinctly will this sys- 
tem of things be apparent; and instead of straight lines, or parts 
of great circles, it will be found that the predominant courses in 
the earth’s features are curves. All these points might be abun- 
dantly illustrated by the groups of the Pacific islands; but we 
omit the details, as the subject will be fully presented in the Re- 
port, by the writer, on the Geology of the Ocean. Suffice it to 
say, that in the Hawaiian range these principles are distinctly 
Amer. Jour. Sci., ii Ser., Vol. III. —May, 1847. 2 386 
Origin of the Grand Outline Features of the Earth. 
represented; so also in the Samoan group, the Kingsmills, the 
Ladrones, and others. 
The citation from Mr. 
Darwin, in the note to p. 
384, exhibits both parallel 
and transverse lines in the 
Galapagos ; and in the Ca- 
naries there are similar 
facts. The position of the 
Azores here given, well il- 
lustrates the subject. The 
main parallel lines are too 
obvious to require partic- 
ular remark, and the trans- 
verse are also apparent. 
A system of curves, on a grand scale, is seen along the east of 
Asia, resembling figure 5. The reader, to appreciate the facts, 
should refer to his map, and the best and largest within reach. 
The first of these curves extends from Kamschatka south by 
the Kuriles to Yeso, and is 1500 miles long ; a second, from Yeso, 
or the island Sanghalian just north, along Niphon to its south- 
west extremity, 900 miles long; a third, from the southwest ex- 
tremity of Niphon, through Kiusiu and other islands, to Loochoo 
and Formosa, about 900 miles long; a fourth, from Formosa, by 
Luzon, Palawan and the western coast of Borneo, 2000 miles 
long. These curves are singularly alike in form and relative 
position. 
These coincidences are facts: accidental, that is, without a 
cause, no one will pretend. The Alaschka Archipelago, at the 
north, seems like a part of the same system; it forms a regular 
curve, 1600 miles long, between Kamschatka and Russian 
America. 
Another corresponding system is apparent in the west coast of 
Asia, though less regular, as its outline is varied by differences 
in the extent of the plains and slopes of the land. But we 
cannot fail to observe the curving lines from Okhotsk south to 
Pekin; from Pekin to Tongquin; and perhaps also from Tong- 
quin to Malacca. The mountain ranges of the eastern side of 
the continent also correspond, as laid down on the best charts. 
The Stanovoi and the Khingan mountains form three great 
curves of similar character, convex in the same direction; and 
the Altai range, farther in the interior, is parallel with the last. 
When the particular islands in the curved lines south of Kam- 
schatka, are laid down with minute accuracy, there is reason to 
believe that each of the curves pointed out, will be found to be 
not a simple curve line throughout, but a compound one, having 
some degree of resemblance either to figure 2, 3, 4, 5 or 6. 
Fig. 7. 
Azores or Western Islands. Origin of the Grand Outline Features of the Earth. 
387 
On a large recent map of the Pacific and East Indies, the range 
of the New Hebrides (K, fig. 8) and the nearly parallel New Cal- 
edonia, (M,) is observed to be continued in the Salomon islands, 
New Ireland and Louisiade group, (I, G, H,) as before stated; 
but the range, we remark farther, is becoming to the westward, 
gradually more east and west in direction, changing from N. 40° 
W. to N. 65° W. The range does not stop here : it is continued 
Fig. 8. 
A. Sumatra; B. Java; C. Sumbawa; D. Ceram; E. 
North New Guinea; F. South New Guinea; G. Admi- 
ralty and New Ireland; H. Louisiade; I. Salomon Group; 
J. Santa Cruz Group; K. New Hebrides; L. Brittania 
Group; M. New*Caledonia; N. Northeast Australia; O. 
North New Zealand. 
in New Guinea, (E, F,) falling off still more towards an east and 
west course ; and the southern division of it, at least, is continued 
farther through Flores, Sumbawa, and Java, (C, B.) and from 
the last island trends northward through Sumatra (A) and the 
Andaman Islands. This is an example of a long curving range ; 
and we may properly connect with it, northern New Zealand, 
northeastern New Holland, and the islands between this coast 
and New Caledonia. Viewing the broad band, we observe the 
whole conforming to one system; and the separate parts, if ana- 
lyzed on a good map, confirm each the same principles. 
Malte Bran mentions the great range of the central Pacific from 
the Marshal Islands by the Samoan to the Austral Islands south 
of Tahiti, which is full 5000 miles long. Leaving the particu- 
lars for another place, we observe only that this great range curves 
north as it goes westward, varying from N. 66° W. to N. 35° W. 
It is also true of mountain ranges that they have this com- 
pound character, though they have seldom been surveyed with 
sufficient care to allow of deriving much accurate knowledge of 
them from maps or descriptions. 
In the valuable work on New South Wales, by Strzelecki, this 
intelligent and laborious traveller mentions and figures the succes- 388 
Origin of the Grand Outline Features of the Earth. 
sive curves, convex westward, which characterize the mountains 
of Eastern Australia, and without reference to any hypothesis, or 
to such a system of things elsewhere. Profs. Rogers in their elab- 
orate papers on the Appalachians, mention that these mountains, 
in their course from Maine to Georgia, are made up of a series of 
great curves, which they describe separately and with detail. We 
shall allude, on a following page, to Dr. Percival’s interesting ob- 
servations on the trap ridges of New England, which sustain the 
same principles in all their detail. Sufficient has been brought 
forward to illustrate the general fact, that the great chains of 
mountains, as well as of islands, are interrupted ranges, consisting 
of overlapping lines, either straight or curved; and that curves 
constitute an essential feature in the system. We have, there- 
fore, but a small part of the truth in the conclusion before stated, 
that there are two prevalent trends in the system of the earth. 
There is still another point to be observed before we are pre- 
pared to draw any conclusions from the facts. Namely:— 
YI. The approximately rectangular intersections of two sys- 
tems of trends wherever they occur together. 
The curving direction of the Java range has been pointed out 
in its course from Sumatra east. Looking again at the map, the 
reader will observe that the coast lines of the large islands north, 
are approximately north and south in direction ; but vary exactly 
with the Java curve. Celebes and Gilolo are north and south 
{ef fig- 8) like western Mindanao; and correspondingly, the 
Java line in the meridian of Celebes, is east and west. The east 
coast of Borneo varies a little to the east of north, and a line 
drawn along it (cd, fig. 8) would meet the Java range at right an- 
gles, or where this range inclines about as many degrees to the 
north of west. The west side of Borneo varies forty degrees to 
the east of north, (ah, fig. 8,) and at the same time the Java 
range, where the line of this side continued would meet it, has a 
like variation to the north of west, not differing even a degree, 
thus making the intersection rectangular. Hence it would seem 
that the shape of Borneo was connected in origin with the trend 
of the Java range ; and not only this, the whole surface covered 
by the islands from Luzon to the Java range, has nearly the same 
shape as Borneo. 
The successive curves on the east coast of Asia, are nearly at 
right angles with one another at their extremities. Thus Niphon 
stands nearly at right angles with the south extremity of the 
Kurile range; so Kiusiu, with the same extremity of the Japan 
range : and also Formosa with that of the Loochoo range. 
In New Zealand, the two systems, as shown by the outline of 
the group, are nearly at right angles. The Tonga range is nearly 
at right angles with the Samoa or Navigators. Passing by other 
facts in the central Pacific, the Galapagos present the same rect- Origin of the Grand Outline Features of the Earth. 
389 
angularity of the two systems. The line of active vents in 
Mexico and that of the great chain are at right angles, as stated 
by Humboldt, and the former is parallel with Cuba. The Cana- 
ries present the same facts as the Galapagos. 
We often find parts of a chain at right angles with the rest, as 
illustrated in figure 6. In the chain of lakes from Lake Erie to 
Bear, which has the northeast course, several of the lakes them- 
selves are oblong across this course. This is the case with the 
parts of Bear lake, with Slave lake, Athabasca, and the northwest 
shore of Superior; and the whole line is at right angles with the 
line of the St. Lawrence, Ontario, and Erie. Indeed such facts 
are closely connected with those first stated with regard to the 
relation of the two prevalent trends of the globe, the northeast 
and northwest, and have long been recognised. 
We have not alluded to a highly important branch of the sub- 
ject :—the direction of cleavage joints in rocks. It has often been 
observed that there is a general correspondence between them 
and the direction of the mountain ranges of a region. Necker 
has presented a great variety of facts on this subject, showing 
the prevalence of northeast and northwest lines in Europe and 
elsewhere.* Professor Sedgwick, in 1831, stated the law of par- 
allelism.! De la Beche mentions the same in Devon and some 
other parts of England, where north-northwest and a transverse 
direction are the common Phillips observes that in 
Yorkshire, fifty-five out of eighty-nine of the cleavage joints ob- 
served by him, were between northwest and north, and twenty- 
eight were at right angles with these ; only six were anomalous. 
The same facts have been remarked by other English geologists. 
Fitton has presented similar facts from Australia. Mr. Darwin 
in his work on South America, gives various facts showing that 
the principle holds west of the Andes, that the cleavage joints 
are in general parallel to the mountain range. It is also true of 
the United States, east of the Appalachians. We observe there- 
fore that the question with regard to the cause of this structure 
is intimately connected with that of the origin of mountains. 
This survey of the geological features of our globe leads to 
several important conclusions. 
A. That the earth has a strongly marked physiognomy, or a 
system in its grand outlines. 
B. That throughout this system, northwest and northeast lines 
are every where prevalent. 
C. That these strongly drawn lines are usually curved in- 
stead of conforming to the direction of a great circle; and 
* Bibliotheque Univ., de Geneve, xliii, 166. 1830. 
t Trans. Geol. Soc., London, ii ser., iii, 68. 
t Geol. Rep. on Cornwall, Devon, and W. Somerset, 8vo, London, 1839. 
§ Sketch of the Geology of Australia, Phil. Mag., Ixviii, 135. 390 
Origin of the Grand Outline Features of the Earth. 
whether curved or straight, consist of a series of subordinate 
parts ; these parts having often a different direction from the 
line of the range. 
D. That the lines, even when curving, cross or meet any 
transverse lines very nearly or quite at a right angle, the one 
dependent on the other or varying with it. Consequently— 
E. That the same grand chain may vary even sixty degrees 
or more in its course, and hence the trend of a ridge is no inde- 
pendent evidence of its age. Thus a northwest course may 
gradually change to an east and west, as in the great Java range 
from New Hebrides to Java, and thence become northwest again, 
as in Sumatra:—a west-northwest range may change to north- 
northwest, as in the great Pacific chain from the Society Islands 
to the northernmost of the Marshall group. A north-northeast 
range may change through northeast, to east and west; and also 
a north and south range may go through the same changes, as 
shown along the east coast of Asia and elsewhere.* Just west 
of New Guinea the east and west line is a little north of east in 
Timor. Consequently, while northeast and northwest lines are 
on the whole most common, there are other courses to be consid- 
ered, and all are so dependent that they evidently must have a 
common explanation. 
Causes of the Earth’s Features. 
The direction of mountain chains is universally attributed to 
the courses of former fissures in the earth’s surface ; and as islands 
come under the same head, and coast lines are mostly dependent on 
the ranges of heights adjoining, the question before us is reduced 
to this :—What can have occasioned such ranges of fissures, with 
their several peculiarities; their composite character, general uni- 
formity of direction, curves, irregularities, and usual rectangular 
intersections ? 
Peculiarities of Fissures.—Before proceeding farther, it is im- 
portant to understand the general character of fissures; and we 
present a case to the point from the map accompanying the elab- 
orate Report of Dr. J. G. Percival on the Geology of Connecti- 
cut ;f—a work of vast labor, and of minute and cautious research, 
by one of the ablest men of America. Dr. Percival has afforded 
us a key to this subject, of the highest value, by the results of 
his investigations among the trap dikes of New England. The 
chart annexed is a small section of the map near its centre, show- 
ing the positions of the dikes or trap ridges. They commence 
twelve miles south of the portion here given, in the vicinity of 
New Haven bay, and extend northward into Massachusetts, and 
beyond into New Hampshire. 
* The same principle is recognized by the Professors Rogers, in view of the facts 
observed by them in the Appalachians. Trans. Assoc. Amer. Geol., 1840-42,p. 540. 
t Report, &c., by James G. Percival, 495 pp. 8vo., New Haven, 1842. Origin of the Grand Outline Features of the Earth. 
391 
The dikes (courses of fissures) have the following character- 
istics :— 
1. A general uniformity of direction. 
2. A situation in several parallel ranges. 
3. An interrupted character, and a frequent advancing or re- 
ceding in the successive parts of a line, or an overlapping of the 
extremities, as in figures 1, 2, 4, 5, constituting what Dr. Perci- 
val has well designated “ advancing,” “ receding” or “ contin- 
ued” series. 
4. Curved lines; some simple, others composed of several 
straight lines, and others of subordinate curves. 
5. Various irregularities in the lines, and deviations from par- 
allelisms, although belonging to the same general system. 392 
Origin of the Grand Outline Features of the Earth. 
These peculiarities, as laid down and described by Dr. Perci- 
val, are confirmed by the writer’s observations on dikes elsewhere 
and by fissures in volcanic regions. 
We have in the above, an example on a comparatively small 
scale, of the general characters of fissures; and we find that it is 
an almost exact representation of the facts presented by the 
prominent features of the earth. The coincidences confirm the 
view that the ranges of mountains and islands correspond to 
ranges of fissures, and also illustrate their subordinate peculiarities. 
The formation of many associated fissures, instead of a single 
rent, is the natural result from the general character of the mate- 
rial ruptured, and the manner in which the force must act. 
Causes of the general uniformity, and composite character of 
ranges.—The most important point with relation to the ranges, 
is their general uniformity over the globe. These great systems 
of parallelisms must have arisen from the ruptures taking place 
in certain directions rather than in others, and the cause lies either 
in the forces causing ruptures solely, or in them in connection 
with the nature or structure of the earth’s crust: and if the lat- 
ter, the structure must be coextensive with the world, as the 
facts have no narrower limits. The 
annexed figure illustrates as we 
have stated a common character of 
ranges. To produce such a result must there not be two distinct 
causes, a force to rupture, and a structure to determine the direc- 
tion of the lines ? Were there no structure, the force should have 
produced a fissure in the general direction of the dotted line A B. 
But instead of this effect it produced a series of parallel lines ob- 
lique with this course. We can conceive of such a systematic 
result only on the ground that there is a tendency to fracture in 
a certain direction; and the force is applied obliquely to the lines 
of easiest fracture. Such ranges, as that of the Hawaiian for 
example, may exceed a thousand miles in length : and the great 
central range of the Pacific, 5000 miles long, illustrates the same 
point, proving the existence of directions of easiest fracture in 
the very nature of the earth’s crust. The fact of such a structure 
has been suggested by Necker,* De la Beche,f BoueJ and other 
geologists. 
The nature of this structure, or the cause of this tendency to 
break in two directions, is a difficult subject of inquiry. The 
material of the crust, to which we naturally look for an explana- 
tion, is crystallized rock, for all igneous rocks are crystalline in 
their nature on cooling; and we observe that such rocks often 
break most easily, in certain directions dependent on the crys- 
tallization of some one of the included minerals or the position 
Fig. 10. 
* Bib. Univ. de Geneve, xliii, 1833, 180. 
t Geol. Report on Cornwall, Devon and W. Somerset, p. 281. 
+ Bull. Soc. Geol. de France, xiv, p. 439, 1843, and ii Ser., i, pp. 353, 355. 1844. Origin of the Grand Outline Features of the Earth. 
393 
of crystalline grains ; though there may be also other independent 
lines of fracture. There is abundant evidence of a uniformity of 
cleavage direction in the rocks of the surface over large areas, as 
already explained. Such a cause would have acted more uni- 
formly at the first cooling of the surface, when from the previous 
free liquidity, the material was more uniform in character than at 
any time afterward: and even though the material were different in 
different parts, it matters little, since feldspar is common in almost 
every igneous rock, and is a frequent source of cleavage in two 
directions at right angles with one another, independently of 
the foliation due to mica and hornblend when either of these 
minerals are present. 
M. Necker, in the article already alluded to, suggests that the 
trends of mountains, coast lines, and the strike of strata, coincide 
with magnetic curves. The same cause is appealed to by Boase* 
and De la Beche, on the ground that the electrical currents trav- 
ersing the globe may influence the polar forces of crystallization. 
It has since been demonstrated by Mr. R. Hunt that the direction 
of crystallization is influenced by magnetism,! and R. W. Fox 
had before shown the action of electrical forces in determining 
the direction of The magnetic chart of the world, 
by Captain Sabine in the sixth Report of the British Association, 
exhibits some striking coincidences throughout the Pacific; the 
exact trend of both the Hawaiian and Samoan groups and also 
the east and west courses in the East Indies and in the West In- 
dies, correspond with lines on his chart. The exceptions are many 
and look insurmountable ; but they are to some extent removed 
by a knowledge of other sources of influence. It should also 
be remembered that lines of magnetic intensity, as Brewster has 
shown, correspond nearly with isothermal lines ; and the two 
agencies, heat and magnetism, must therefore have acted in some 
degree together, at all periods. <§> 
Hopkins in his able “ Researches on Physical Geology,”|| 
(1835,) attributes the regularity of joints in rocks to the mechan- 
ical action of an elevating force, and he establishes a perfect uni- 
* Treatise on Primary Geology, by H. S. Boase, M. D.; 8vo, London, 1834, and 
L. and E. Phil. Mag., and Jour, of Sci., ix,4; x, 14. 
t Phil. Mag., Jan., 1846, p. 1; Amer. Jour. Sci., ii Ser., ii, 116. 
t Report of the Polytechnic Society of Cornwall, for 1837, pp. 20,21 and 68, 69. 
§ M. Boue observes:—Sans avoir besoin de supposer que la terre ait ete prim- 
itivement un cristal, it suffit de lui accorder un certain ordre dans ses irrdgularitis 
de surface en harmonie avec son mode dc refroidissement, ses forces int6rieurcs et les 
forces centrifuge et centripete. Ceci admis, on en doit deduire necessairement que 
les premieres mers ont occupfe les parties du spheroide les plus accidentees, cer- 
taines grandes chaines offrant encore les indices de ces formes regulieres, ou pseu- 
do-reguli&res; et puisque ces series de montagnes constituent l’ossature des conti- 
nents, et determinent leur figure, on voit de nouveau combien la similitude des 
continents eolaire l'etude pour ainsi dire cristallographique du noyau terrestre. Bull, 
de la Soc. Geol. de France, i, ii Ser., 355. 
II Trans. Camb. Phil. Soc., vii, 1. 
Amer. Jour. Sci., ii Ser., Vol. III.—May, 1847. 3 394 
Origin of the Grand Outline Features of the Earth. 
formity between the facts and the necessary effects of this cause. 
Mr. Darwin adopting essentially the same view in his remarks 
on the parallel relation of the planes of cleavage in western 
South America to the axis of the Andes range, explains the uni- 
formity by supposing the mass to have been subjected to tension 
unequal in different planes, arising from the elevation of the 
mountains.* Mr. Sedgwick in his valuable memoir u On the 
Structure of Large Mineral Masses,” (1835,) also appeals to ten- 
sion as the cause, and supposes that this tension may arise from 
the contraction attending solidification.! If tension be the prox- 
imate cause, the various facts require that both sources of it, the 
mechanical and that of refrigeration, be equally appealed to.J 
With reference to this subject, it should be considered that if 
curves of magnetic intensity are approximately isothermal lines, 
they must have been lines of equal cooling, and consequently 
lines of equal tension. This cause would then cooperate with 
the electrical, and might aid in producing the general uniformity 
of trend, which could not proceed from contraction alone. Act- 
ing during the period of early cooling, its effects should therefore 
have been universal: and through subsequent ages, the cooling or 
crystallization, beneath the crust, making still slower progress, 
(inconceivably slow,) would have continued to be governed by 
the same cause, liable to those modifications that isodynamic 
lines have undergone. But a perfect correspondence in the sur- 
face plutonic rocks, with the structure of the crust, should not be 
expected, since the lines of tension, determining the structure of 
the former, must have depended somewhat on the direction of 
the force producing fissures and mountain elevations. 
The bare possibility that the earth’s axis of rotation has been 
at any time changed, suggests that a cause may have cooperated 
in these results, whose influence cannot be fully estimated. Yet 
if magnetism has been a cause of structure, the coincidence of 
trend with existing curves of magnetic intensity tends to prove 
that such a change of axis has not taken place. It would be a 
grand result for geology if the science should settle this debated 
point. The coincidence of the magnetic curves with the trends 
in the central Pacific and north of New Holland, and also in the 
West Indies, is so close, that we have reason to suspect that the 
two facts are some way mutually dependent and have always been 
so. Still if the earth’s axis may have changed, it is possible that 
the trends may once have had a direction that could have been 
* Darwin on South America. 8vo, London, 1846. p. 163. 
1 Trans. Geol. Soc., London, ii Ser.,iii, 480, March, 1835. 
t Tidal movements in the fluids during incipient cooling might be a source of 
tension, transverse to the line of motion. And a gradual change in the oblateness 
of the globe would be another source of tension. But it does not appear that the 
existing system would correspond with the possible effects of either of these 
causes. Origin of the Grand Outline Features of the Earth. 
395 
mainly caused by tension consequent on a diminution of the 
earth’s oblateness. 
Whatever the origin, there can be no doubt of the fact, that a 
kind of cleavage structure, or, at least, a capability of fracturing 
most easily in two directions, was given the crust during its form- 
ation, and that such a structure has influenced the direction of 
the lines of fissures that have since taken place. And while there 
is evidence of this structure, there is proof that the rupturing 
force often acted obliquely to the planes of easiest fracture, caus- 
ing deviations from straight lines in the long ranges. The next 
question is with reference to this rupturing force. 
Contraction is a known dynamical cause that must have begun 
with the beginning of refrigeration; and it is hence essential to 
consider how far it meets the various facts in view. In the the- 
ory of mountain ranges, by Elie de Beaumont, this agency is ap- 
pealed to; we believe as confidently in its efficiency, though led 
by the facts to somewhat different results. 
The effects of contraction have been illustrated elsewhere in 
this volume. A prime feature in the operation of cooling, influ- 
encing all the results proceeding from it, depends on the tendency 
of heat to spread itself circularly, or to diminish circularly, around 
a centre. This cause gives a circular form to pools of lava, and 
they retain this form as they cool. The great crater areas of the 
moon, several hundred miles in diameter, illustrate it; and this 
size is no necessary limit to their extent. In a cooling globe 
there would therefore be necessarily such vast circular or elliptical 
areas. Here then we perceive a cause modifying all the results 
of cooling; and we observe that throughout all ages there must 
therefore have been some reference to such circular or elliptical 
areas in contraction ; and especially, to aggregations of such areas, 
which also would be more or less curvilinear in outline, and 
would act as a whole in the progressive subsidence. 
The force of tension in the crust from contraction beneath, 
is exerted to a great extent horizontally; and in a subsiding area, 
the direction would be nearly radial, or from the centre outward. 
This cause then should generally act obliquely to the lines of 
structure, though sometimes coincident with them. If the ten- 
sion cause ruptures, the rents should follow the lines of cleavage 
structure in the earth, in case the direction of the force corres- 
ponded ; otherwise a series of rents should result having a direc- 
tion of range different from the direction of the line of structure. 
The peculiarities of fissures, which have been explained, the 
“receding,” or “advancing,” or “continued” series of parallel 
courses, and the curved directions, are therefore necessary effects 
of the cause appealed to. Curved as well as straight ranges, 
ought therefore to characterize the grand features of the globe. 
The important generalizations of Mr. Hopkins with regard to 
the direction of fractures and the necessary dependence of two 396 
Origin of the Grand Outline Features of the Earth. 
transverse lines in an elliptical area under a state of tension, not 
only remove any difficulty arising from the existence of two 
transverse systems and their rectangular intersections, but actu- 
ally require this result.* 
Areas of non-contraction or of comparative slow contraction, 
should modify the direction of the ranges of fissures formed in a 
surrounding region where more rapid contraction is going on. 
Also, the interference of two contracting areas would produce 
irregularities; still wider effects would proceed from more ex- 
tended combinations, such as have produced the oceanic depres- 
sions and the continental areas. Thus the continents which 
were early free from firesf have generally experienced the tension 
along their borders ; and fissures and mountain ranges, frequently 
several in parallel series, have been formed, whose main courses 
are a resultant between the direction of the planes of cleavage 
structure in the earth and the action of the force of tension arising 
from the contraction going on over the oceanic areas.;); Causes 
of certain irregularities in mountain ranges were mentioned on 
page 185 of the last number of this Journal, and these discussions 
afford a more extended view of the action of these causes. The 
principles explained in the paper just referred to, have here their 
full application. 
Any other cause besides contraction, occasioning elevations or 
subsidences would produce the same general result as regards di- 
rection of lines of fissures ; but we know no other cause of prob- 
able operation that would be so related to elliptical or circular 
areas; and thus none but this cause appears to satisfy the con- 
ditions presented in the frequent curvilinear forms of ranges. 
The positions of some great contracting areas may be dis- 
tinguished over the oceans, from the curving lines which enclose 
them. The great Pacific range of lands, from the Marshall Isl- 
ands to the Society group, 5000 miles in length, has been de- 
scribed as convex to the southwest, while the line of the Ha- 
waiian range, 2000 miles long, is nearly straight. May not this 
part of the ocean have been one of the large compound contract- 
ing areas, and a line from Pitcairn’s, in lat. 25° S., long. 130° W., 
to northern Japan the course of its axis ? Using the registers of 
* The mathematical deductions of Mr. Hopkins were made with special refer- 
ence to the elevation of the Wealden, though brought out so as to be of general 
applies ion. He lays down the facts, that in “ districts where faults abound, two 
distinct systems are usually found, in each of which the faults approximate to 
parallelism with each other;” and that “ the common direction of one of these 
systems is approximately perpendicular to that of the other;” and he establishes 
the necessary dependence of these transverse systems by calculation. 
The mode of producing the tension required by fracture is different in the fore- 
going explanations, from what is assumed by Mr. Hopkins; but it does not appear 
to alter the general results; and it is believed to set aside some objections urged by 
Dr. Boase to the conclusions of Mr. Hopkins. See L. and E. Phil. Mag., and J. 
of Sc., ix, 4, 171,368, and x, 14. 1836. 
i See this Journal, ii, 132, and iii, 181. t Ibid, iii, 98,181. Origin of the Grand Outline Features of the Earth. 
397 
subsidence, so happily distinguished and brought forward by Mr. 
Darwin, the Coral Islands, we have evidence that an elliptical 
area with the same line for its axis was subsiding even as late as 
since the tertiary epoch. The very region therefore which bears 
evidence of having been the original great elliptical area of con- 
traction for the Pacific, on which the courses of the islands were 
in part dependent for their direction, was also undergoing con- 
traction till within a late period; and we know not that some 
parts about the Northern Carolines—the nearest to the centre of 
the area—may not still be contracting as there is some evidence 
of it, which the writer will elsewhere present. The transverse 
line including New Zealand, the Kermadec and Tonga Islands, 
crossing the other systems nearly at right angles, would pass in its 
course northward the Samoan and Hawaiian Islands, and also 
some smaller groups intermediate. 
The position of a large area undergoing little contraction com- 
pared with the region around, is before us in New Holland, as is 
evident from the absence from this semi-continent of volcanoes 
or their remains. Such an area would occasion a tension acting 
to some extent circularly around it; and which might determine 
the courses of ranges in its vicinity. The ranges of islands from 
New Zealand by the New Hebrides to New Guinea and Java, is 
just such a concentric range, as the view would seem to require. 
Borneo is another vast region without volcanic traces over its in- 
terior, and may have influenced the upward trend exemplified in 
Sumatra. However this may be, the cause brought forward— 
large isolated areas of comparatively slight contraction,—must 
have had their influence in determining the direction of lines of 
tension or of forces causing rupture. 
M. Boue remarks that the trends in the tropics in general ap- 
proach a parallelism with the equator, and he attributes the sup- 
posed fact to the centrifugal force of rotation. It holds true to 
a considerable extent. There are however so many exceptions 
that we may perhaps doubt whether the fact is sufficiently gen- 
eral for so general a cause. 
The conclusions which appear to flow from the facts that have 
been presented, are as follows:— 
That the general direction and uniformity of the grand outline 
features of the globe may be in a great degree the simple effects 
of the earth’s cooling: this operation resulting in (1) solidifica- 
tion, and under the circumstances, whatever they were, an attend- 
ant jointed structure or courses of easiest fracture, in two direc- 
tions at right angles nearly with one another, both varying to- 
gether according to the rates of cooling in different parts ;—and 
(2) occasioning tension in the crust through the contraction going 
on beneath, with some relation to circular areas but especially to 
large compound areas, which tension caused ruptures conforming 398 
Origin of the Grand Outline Features of the Earth. 
or not to the lines of jointed structure according as the force of 
tension acted in accordance with this structure or obliquely to it. 
(3) The age of mountains cannot therefore be determined neces- 
sarily by their courses ; a different direction in a particular region 
in different ages is not improbable, since the same contracting area 
might exert its horizontal force in somewhat different directions 
at different epochs, or other such areas might cooperate, and exert 
a modifying influence ; and at the same time, an identity of di- 
rection for different ages was to have been expected. 
From the facts before us it may be inferred that the great vol- 
canic band which is drawn by von Buch in the East Indies, in 
the shape of the letter U, from Sumatra and Java around by the 
Philippines, gives an incorrect view of the volcanic system in 
that part of the world. Much the larger part of the Philippines 
consists of primary and secondary rocks instead of volcanic, and 
in Luzon, the southern volcanic portion corresponds nearly to 
the west-northwest trends of the Pacific. The volcanic line of 
Sumatra and Java, including also the islands farther east, belongs, 
as we have shown, to the system of the Pacific, to the north- 
westward courses which characterize nearly all the groups of that 
ocean ; and it is continued east by a volcanic line along North New 
Guinea to the New Hebrides. Although an erroneous impression 
may therefore be conveyed by the chart of von Buch, it presents 
properly the fact intended to be illustrated by its distinguished au- 
thor, that volcanoes prevail along the track laid down. The band 
seems like a grand volcanic border to the Asiatic continent, stretch- 
ing from the vicinity of northern New Holland, (and we may say 
from New Zealand,) to Kamschatka; and it curves around into 
the great American range through the Alaschka Archipelago. 
It is a fact of no little interest that the Pacific Ocean should 
thus be nearly encircled with volcanoes, active or extinct, as well 
as by high mountains, while the sides of the narrow Atlantic 
and Indian Oceans have comparatively few traces of such fires; 
and it tends to confirm the opinion thrown out as to the agency 
by which the deep ruptures and elevated ranges of the globe have 
been produced.* 
This subject properly forms an introduction to the preceding 
papers on the earth’s contraction, in this volume ;f and the whole 
is offered as a simple view of the earth’s dynamics, the suffi- 
ciency of which is to be tested by future discoveries. 
Note.—On page 188, it is implied that LycII and Poisson admitted the former 
fluidity of our globe ; whereas the writer simply intended to acknowledge that 
these authors first brought forward the argument that a crust could not form upon 
a globe while it was in a state of free liquidity. See Lyell’s Principles, ii, 439, 
where Cordier is mentioned as having presented the argument subsequently to 
himself. 
* See this volume, pp. 96, 98. 181, 185,186. t Ibid, pp. 94, 176.